or Connect
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Tube amp with cheap transformers 'useless' ?
New Posts  All Forums:Forum Nav:

Tube amp with cheap transformers 'useless' ? - Page 2

post #16 of 106

I was thinking of the traditional OTL designs that parallel many tubes together. The Zana Deux you mention uses 6C33C-B power triodes to get that current output and costs $2200.00. Those kind of tubes were used as voltage regulators and power amps for servos, etc by the Russian military. So yeah, I stand corrected. You can get that much power out of a single tube. But, for that price you can get a good transformer coupled amp too.


When I mention that cheap iron can be quite good, I don't mean crappy transformers. There are transformers out there that are really good quality but only less expensive because they were made in quantity for another purpose, say a 70V speaker transformer or an SE guitar amp transformer. I've done experiments with 70V speaker transformers and have found them to work quite well in this, as well as other applications with the correct topology. Low distortion, super low output impedance (1 to 2 ohms) and flat response. Sounds great, too. I've built several.


If you go the traditional route and buy a purpose built output transformer from an 'audiophile approved' vendor, then expect to pay. If you want OTL and need to drive current hungry phones you need real power, which means big ass tubes or many tubes in parallel, so again, expect to pay.


The Decware amp mentioned as well as the Little Dot use a 'totem pole' or SEPP output. These can drive low impedances but not in the 16 to 32 ohm realm with any kind of current. The Decware claims an output impedance of around 60 ohms. Best to pair this with something in the 150 to 600 ohm range. The Little Dot's spec show that the power output diminishes the lower in impedance you go and the Decware would be the same due to similar topology. I haven't heard either of these amps; maybe they do a good job on low Z phones, but the numbers have me doubting they would do much for bass heavy music on inefficient headphones.


But, the original point about price seems to be true, those two examples are relatively inexpensive since they don't use trannys.

Edited by elliottstudio - 8/6/12 at 9:36pm
post #17 of 106
$1500 is just too much. Even with 100% markup.
(Output) transformers perform better (less distortion) when lightly loaded. How much lighter a load are you going to get than headphones? Admittedly interstage transformers have even lighter loading, but headphones are a light load. You don't need anything more than a low power (speaker) transformer, run at fractional power.
You can buy 2 Hammond 10W push-pull Hi-Fi output transformers for 125$. Even if you go single ended, 2*10W transformers will still only be ~$200. These are truly overkill in a headphone amp. You can probably use the lower grade guitar amp style transformers considering the light loading.
I'd expect to be able to build a pretty good (beyond reproach) tube headphone amp (that'll incidentally drive speakers), including buying transformers, for $400 at the outside. A manufacturer can expect some economies of scale when purchasing.
Still, you can probably build a couple of good quality 200WPC solid state amps for the same price.
Alternately, there is the modified Morgan Jones OTL design using 6N1P Russian tubes.
- drives reasonable current into 64 ohms, probably OK with pretty much anything since lower impedance phones tend to be efficient.
post #18 of 106

any small manufacturing company serious about meeting overhead, pay employees, existing for the long term in a 1st world nation is charging ~5x parts cost


expectations are warped if you consider mass market items with millions in volume as your pricing guideline

post #19 of 106

400% markup? You may be right, but I don't think so.


JDS labs sell a complete, assembled O2 for $144, which I reckon to be about 100% markup. Of course they may go broke selling them, but we'll see...



post #20 of 106

there can be exceptions - one man kitchen table operations - the owner/operator is unlikely to be paying themselves a competitive wage


or your "manufacturing" can just be a import business - less multiple is possible if you buy finished assembles and at most silkscreen your name on them - but you have to buy from a cheap labor country and/or have sales volume numbers higher than most headphone tube amp models


HeadRoom originally claimed to try to hold to a 3x multiple - don't know if that's still working for them - but they also have the headphone resale income - which cross subsidizes which?

post #21 of 106
Thread Starter 

I'm with jcx on this - makes you wonder what the 'invoice cost' for Grant Fidelity's tube gear must be. 

post #22 of 106

I don't doubt that there are companies employing even bigger markups than x5. SS headphone amps @ $600+ spring immediately to mind.


There's the Grado RA-1 @ $450 for no more than $20 worth of parts, most of that going on the wooden box.


Items such as this being commonplace in the market degrades the normal sensibilities of buyers.


I neither to work for nor buy from such. I don't approve of and I am not an apologist for them.


I don't believe such markups are necessary to survive in business. With items retailing at hundreds of dollars, volume sales are not a necessity. If I sold my putative $400 amp at your markup ($2000) the excess would be $1600 while the production costs would barely exceed that of an O2. You don't need to sell many of those in a year as a one-man business to survive, and you're hardly going to be run off your feet at, say, one-a-fortnight ($38400 PA before tax).


I accept that I have oversimplified here, but $1500 is too much for a tube headphone amp, too much for any headphone amp, in my opinion.




'One cannot truly blame them for this; one can only despise them' Frank Herbert in Dune

post #23 of 106
This isn't something that requires any speculation. Any manufacturer of any significance needs to get at least 60% gross margins on the products it sells. And typical reseller margins are 40%. So if you make something that costs $1000 to the end user, the reseller buys it for $600, and the COGS should be about $240. So yes, about 4x COGS is what the end user price will typically be, although you can often get some discount off the list price from the reseller.

Unless you are truly just a one or two man operation, you need to have this kind of markup to have any chance of having a successful business. Remember, cost of goods doesn't include R&D costs, marketing costs, selling costs, or anything else.
Edited by Skylab - 8/7/12 at 6:06pm
post #24 of 106
Originally Posted by wakibaki View Post

I don't believe such markups are necessary to survive in business. 



You have no idea what you are talking about. 

In pretty much this whole thread. 

post #25 of 106

As a one man business myself, I have put countless hours into research, experimentation, designing, prototyping, drafting etc. before I even commit a design to sheet-metal. None of those hours are charged into the retail price of my products, I did it as a hobby and labor of love. I sell direct to keep costs low and don't advertise. My marketing skills suck and I'm not selling lots of stuff, but I don't feel like I could raise the price of my products any higher so I can't justify the BIG $$ that would go into marketing. And still I have to charge what many would consider a high price. Sheet metal casework is expensive, so is engineered shipping boxes and packing materials, not to mention the parts and wire that go into the product itself. So if I sell a preamp for $3000.00, I make a fraction of that after all the expenses and paying myself are figured in. I couldn't imagine hiring anyone to help!


Of course the larger manufacturers sell lots of gear and can make up some of the initial outlay in volume. As well as pay the salaries of their employees, taxes, leases, overhead, marketing, etc. I wonder how some of the American 'affordable' brands do it without outsourcing to China. I'm really glad they are successful and think they're products are a good value, especially the ones that do go through distribution and have to pay all the middle men.


Me- I haven't given up my day job.... yet.

Edited by elliottstudio - 8/7/12 at 7:42pm
post #26 of 106
Thread Starter 

Someone mentioned taxes - I dont know about the US or UK, but in Oz starting a business is an open invitation to Federal, State and Local governments to put their hands out. So many hidden costs that wage slaves like myself dont even have to think about, thankfully. 

post #27 of 106
Originally Posted by nikongod View Post


You have no idea what you are talking about. 

In pretty much this whole thread. 


What? JDS labs aren't selling the O2 at $144?


Like I said earlier, degraded sensibilities. You guys are all part of a rip-off culture.



post #28 of 106
Thread Starter 

I categorically reject that. Would I buy high-end audio if it were only 2-3 times the price of DiY ? Your damned right I would. Is that going to happen anytime soon ? No, it isnt. Whether I like it or not, gold costs more than copper - both are valuable metals, but we have simply assigned more value to gold than copper. Where it gets nasty is when someone tries to turn a lump of coal into a diamond. That is what's known as 'fools gold' - some of the more extreme objectivists would have us believe that its ALL fools gold, and I refuse to accept that. Life is fairly short and brutal without aspirations. 

post #29 of 106





more relevant to quality of low cost tube output transformers is the cost of Nickel = only "an arm or a leg" and Cobalt = "need to sell off major organs"


so "low cost" xmfr are pretty much going to use grain oriented Si steel laminanation - giving the possibilty of low level hysterisis distortion being problematic

Edited by jcx - 8/9/12 at 11:07am
post #30 of 106

Yes, and because of that high cost nickel and cobalt are rarely used in any transformer larger than a mic or line input or line output transformer. You could use nickel for a headphone amp output transformer, which would have a relatively small frame due to the low currents involved. I'm sure some manufacturers do, however I doubt any low level hysteresis distortion is even audible; a good grain oriented steel core will not distort until it is overloaded or saturated with DC running through it, such as a SE amp. Parallel feed SE takes care of that. I've run distortion analysis of a tube line stage with a transformer connected and without. Other than the change in level due to the turns ratio there was no difference in the distortion level or spectrum. The distortion generated by the tube stage itself will greatly swamp out any residual transformer induced distortion. Now, when you go up in power things do change.

Edited by elliottstudio - 8/9/12 at 4:59pm
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Tube amp with cheap transformers 'useless' ?